Response allocation in a rapid-acquisition concurrent-chains procedure: effects of overall terminal-link duration.

Four pigeons were exposed to a concurrent-chains procedure in which the initial link was constant (variable-interval 10s) while the terminal-link delays changed across sessions according to a pseudorandom binary series. In the short condition, the left terminal link was a fixed-interval 8s and the right terminal link was either fixed-interval 4s or fixed-interval 16s; in the long condition, the left terminal link was fixed-interval 16s and the right terminal link was either fixed-interval 8s or fixed-interval 32s. Multiple regression showed that response allocation was sensitive to the current session immediacy ratio, with greater sensitivity in the long than the short condition. This replicates the terminal-link effect [MacEwen, D., 1972. The effects of terminal-link fixed-interval and variable-interval schedules on responding under concurrent-chained schedules. J. Exp. Anal. Behav. 18, 253-261] in a pseudorandom binary series design. Grace and McLean's [Grace, R.C., McLean, A., 2006. Rapid acquisition in concurrent chains: evidence for a decision model. J. Exp. Anal. Behav. 85, 181-202] decision model provided a good description of the data and predicted the terminal-link effect. Estimates of the criterion parameter increased in the long condition but less than proportionally, and were close to the expected values based on the programmed schedule durations, as predicted by the decision model. These results provide additional support for the decision model as an account of choice in concurrent chains.